Hugging Face's logo

IrwinD
/
log_sage_reward_model

Text Classification
Transformers
Safetensors
distilbert
trl
reward-trainer
Generated from Trainer
Eval Results
Model card Files
xet
Community
IrwinD commited on
Commit
8cc4fc1
·
verified ·
1 Parent(s): 99ccfee

End of training

Browse files
Files changed (3) hide show
  1. README.md +29 -44
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -34,7 +34,7 @@ should probably proofread and complete it, then remove this comment. -->
34
 
35
  This model is a fine-tuned version of [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) on the hdfs_rlhf_log_summary_dataset dataset.
36
  It achieves the following results on the evaluation set:
37
- - Loss: 0.4346
38
  - Accuracy: 1.0
39
 
40
  ## Model description
@@ -55,11 +55,11 @@ More information needed
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 1.41e-05
58
- - train_batch_size: 6
59
- - eval_batch_size: 24
60
  - seed: 42
61
  - gradient_accumulation_steps: 16
62
- - total_train_batch_size: 96
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
  - num_epochs: 40
@@ -68,46 +68,31 @@ The following hyperparameters were used during training:
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
- | No log | 1.0 | 1 | 0.6931 | 0.8 |
72
- | No log | 2.0 | 2 | 0.6932 | 0.8 |
73
- | No log | 3.0 | 3 | 0.6927 | 0.8 |
74
- | No log | 4.0 | 4 | 0.6916 | 1.0 |
75
- | No log | 5.0 | 5 | 0.6895 | 1.0 |
76
- | No log | 6.0 | 6 | 0.6861 | 1.0 |
77
- | No log | 7.0 | 7 | 0.6800 | 1.0 |
78
- | No log | 8.0 | 8 | 0.6720 | 1.0 |
79
- | No log | 9.0 | 9 | 0.6613 | 1.0 |
80
- | 0.3371 | 10.0 | 10 | 0.6489 | 1.0 |
81
- | 0.3371 | 11.0 | 11 | 0.6380 | 1.0 |
82
- | 0.3371 | 12.0 | 12 | 0.6289 | 1.0 |
83
- | 0.3371 | 13.0 | 13 | 0.6243 | 1.0 |
84
- | 0.3371 | 14.0 | 14 | 0.6158 | 1.0 |
85
- | 0.3371 | 15.0 | 15 | 0.6049 | 1.0 |
86
- | 0.3371 | 16.0 | 16 | 0.5933 | 1.0 |
87
- | 0.3371 | 17.0 | 17 | 0.5812 | 1.0 |
88
- | 0.3371 | 18.0 | 18 | 0.5619 | 1.0 |
89
- | 0.3371 | 19.0 | 19 | 0.5455 | 1.0 |
90
- | 0.2662 | 20.0 | 20 | 0.5372 | 1.0 |
91
- | 0.2662 | 21.0 | 21 | 0.5299 | 1.0 |
92
- | 0.2662 | 22.0 | 22 | 0.5194 | 1.0 |
93
- | 0.2662 | 23.0 | 23 | 0.5001 | 1.0 |
94
- | 0.2662 | 24.0 | 24 | 0.4711 | 1.0 |
95
- | 0.2662 | 25.0 | 25 | 0.4529 | 1.0 |
96
- | 0.2662 | 26.0 | 26 | 0.4462 | 1.0 |
97
- | 0.2662 | 27.0 | 27 | 0.4399 | 1.0 |
98
- | 0.2662 | 28.0 | 28 | 0.4361 | 1.0 |
99
- | 0.2662 | 29.0 | 29 | 0.4353 | 1.0 |
100
- | 0.198 | 30.0 | 30 | 0.4391 | 1.0 |
101
- | 0.198 | 31.0 | 31 | 0.4491 | 1.0 |
102
- | 0.198 | 32.0 | 32 | 0.4516 | 1.0 |
103
- | 0.198 | 33.0 | 33 | 0.4486 | 1.0 |
104
- | 0.198 | 34.0 | 34 | 0.4529 | 1.0 |
105
- | 0.198 | 35.0 | 35 | 0.4516 | 1.0 |
106
- | 0.198 | 36.0 | 36 | 0.4518 | 1.0 |
107
- | 0.198 | 37.0 | 37 | 0.4468 | 1.0 |
108
- | 0.198 | 38.0 | 38 | 0.4412 | 1.0 |
109
- | 0.198 | 39.0 | 39 | 0.4369 | 1.0 |
110
- | 0.1619 | 40.0 | 40 | 0.4346 | 1.0 |
111
 
112
 
113
  ### Framework versions
 
34
 
35
  This model is a fine-tuned version of [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) on the hdfs_rlhf_log_summary_dataset dataset.
36
  It achieves the following results on the evaluation set:
37
+ - Loss: 0.3796
38
  - Accuracy: 1.0
39
 
40
  ## Model description
 
55
 
56
  The following hyperparameters were used during training:
57
  - learning_rate: 1.41e-05
58
+ - train_batch_size: 4
59
+ - eval_batch_size: 16
60
  - seed: 42
61
  - gradient_accumulation_steps: 16
62
+ - total_train_batch_size: 64
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
  - num_epochs: 40
 
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
+ | No log | 1.0 | 1 | 0.6922 | 1.0 |
72
+ | No log | 2.0 | 3 | 0.6862 | 1.0 |
73
+ | No log | 3.0 | 5 | 0.6811 | 1.0 |
74
+ | No log | 4.0 | 6 | 0.6785 | 1.0 |
75
+ | No log | 5.0 | 8 | 0.6729 | 1.0 |
76
+ | 0.2836 | 6.0 | 10 | 0.6659 | 1.0 |
77
+ | 0.2836 | 7.0 | 11 | 0.6611 | 1.0 |
78
+ | 0.2836 | 8.0 | 13 | 0.6493 | 1.0 |
79
+ | 0.2836 | 9.0 | 15 | 0.6328 | 1.0 |
80
+ | 0.2836 | 10.0 | 16 | 0.6230 | 1.0 |
81
+ | 0.2836 | 11.0 | 18 | 0.6019 | 1.0 |
82
+ | 0.2714 | 12.0 | 20 | 0.5754 | 1.0 |
83
+ | 0.2714 | 13.0 | 21 | 0.5604 | 1.0 |
84
+ | 0.2714 | 14.0 | 23 | 0.5312 | 1.0 |
85
+ | 0.2714 | 15.0 | 25 | 0.5008 | 1.0 |
86
+ | 0.2714 | 16.0 | 26 | 0.4862 | 1.0 |
87
+ | 0.2714 | 17.0 | 27 | 0.4728 | 1.0 |
88
+ | 0.2714 | 18.0 | 29 | 0.4492 | 1.0 |
89
+ | 0.2576 | 19.0 | 31 | 0.4277 | 1.0 |
90
+ | 0.2576 | 20.0 | 32 | 0.4183 | 1.0 |
91
+ | 0.2576 | 21.0 | 34 | 0.4027 | 1.0 |
92
+ | 0.2576 | 22.0 | 36 | 0.3909 | 1.0 |
93
+ | 0.2576 | 23.0 | 37 | 0.3864 | 1.0 |
94
+ | 0.2576 | 24.0 | 39 | 0.3808 | 1.0 |
95
+ | 0.1866 | 24.73 | 40 | 0.3796 | 1.0 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
 
97
 
98
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cceeeb14facc4673d1030b199e4872b3f9f1e69956fa6b8ea719478a0976987c
3
  size 267829484
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b48603a13dc746f4d38a7fd963c300e8afaf4f7355a39c6577ac554c1ac9ee4b
3
  size 267829484
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:88014a04210725e0d48578a72ffb51a49027581b5feb0077963cbf63ee3be7ae
3
  size 4984
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:604dded5316d49385a7309eb2e39f4dfe7ab7c5f1e4a90a9730b089ec96f6335
3
  size 4984